Dietary Assessment
We recommend choosing one of two self-reported diet measures, either a Food Frequency Questionnaire (FFQ) or a 24 Hour Recall (24HR). Food Frequency Questionnaires are flexible instruments for collecting dietary information that are commonly used in nutritional and medical studies. FFQs include a list of foods/beverages and a range of frequencies of consumption that respondents can select from for each food; they sometimes also include options for serving sizes. 24 Hour Recalls collect detailed data about all foods/beverages consumed in a 24 hour period, and include different prompts to help improve the accuracy of reporting. They traditionally take the form of structured interviews, but increasingly can also be self-administered online.
In terms of diet self-reports, traditional FFQs are our preferred instrument for non-experimental studies when frequency of consumption and/or categorization (vegetarian, omnivore, etc.) are desired outcomes. If the goal of the FFQ is to identify vegetarians or vegans, to limit over-reporting it is wise to first administer a food list, with respondents indicating only whether or not they ever eat the food in question. Following this, it is useful to ask respondents to self-identify with the label their food lists suggests they correspond with (e.g., if they indicated they ate no meat but did eat eggs and/or dairy, the next question would ask them to self-identify as a vegetarian). This is particularly important if respondents will be asked questions about this diet elsewhere in the survey to avoid frustration for those who for whatever reason have a mismatch between the food list and the way they self-identify.
We are less certain about recommending the use of traditional FFQs in experimental studies, particularly where the desired minimum detectable effect risks going undetected within the frequency scale ranges. For experimental studies, we anticipate that a proper 24-HR (e.g. the ASA24) is the best method, but we recognize that this is an involved process and so hope to have more information soon on how other types of self-reported measures (traditional FFQs and its variations) perform in experimental contexts.
FFQs and 24HRs are generally used to measure respondents’ patterns of consumption. However, animal advocates may not be strictly interested in the number of animal products consumed by respondents; they may also be interested in the total number of animals who suffer or die due to respondents’ food-related habits. In addition to asking how many animal products respondents have consumed, animal advocacy Researchers might consider asking how many animal products respondents have served to others or discarded. Of course, the benefits of obtaining additional information must be weighed against the risk of increasing response burden.
Dietary Recalls & Intervention Research
While FFQs are a useful instrument for describing an individual’s usual consumption—especially when gradations of meat consumption are of interest—they may be less suited to measuring dietary change in experimental settings, particularly if smaller changes in consumption are of interest, however they may also bring advantages such as reduced variance.
A FFQ must be sensitive enough to capture relevant changes, which is referred to as its “responsiveness” (Kristal et al., 1994, page 185S). While FFQs are considered to be the most common instrument for dietary intervention research (Thomson et al., 2003, 754-755), we have concerns about the extent to which they can be responsive to small changes in animal product consumption that may be meaningful to the movement.
Notable changes in consumption can be hidden even among larger frequency scales. Take for example the following 9-point scale: never; less than once per MONTH; 1–3 times per MONTH; once per WEEK; 2–4 times per WEEK; 5–6 times per WEEK; once per DAY; 2–3 times per day; and 4 or more times per day. If a subject in an intervention study reduced their consumption of a particular food group from 4 times per week to 2 times per week, this instrument would not be responsive enough to capture this change. Yet, many Researchers and advocates may consider this a meaningful difference given that it represents 100 fewer servings per year. If this was measuring chicken consumption, it would represent the meat of roughly 8 fewer chickens being consumed annually. If the intervention is likely to reach a large group of individuals this would magnify the change further.
Even if the minimum detectable effect is large enough such that it is less likely to go undetected in a study using an FFQ, there remains some concern about the suitability of these instruments for intervention studies. For example, the National Cancer Institute (NCI, an arm of the National Institutes of Health) recommends in their Dietary Assessment Primer “the use of 24HRs [24-hour recalls] over FFQs or screeners when evaluating the effect of an intervention on diet because 24HR data are less prone to overall bias.” A single administration of a 24HR is NCI’s recommended approach over FFQs for a variety of intervention designs, including: “determining change, between two points in time, in mean usual intake of group,” “determining difference between two groups in mean usual intake change,” and “determining difference between two groups in post-intervention mean usual intake.”
24HR, administered in the traditional sense, represent more respondent burden and likely also cost more than FFQs so they are not suitable for all research designs. Below we present two options, which differ on these matters.
Nontraditional FFQ
In their intervention study of the impact of their Facebook ads, Mercy For Animals used a self-reported dietary measure that asked respondents: “In the past two days, how many servings have you had of the following foods? Please give your best guess. One serving of meat or eggs is 3 ounces, about the size of a deck of cards.” Respondents were able to indicate a single response for each listed food group (0, 1, 2, 3, 4, 5, etc.). This approach has the advantage of asking about a smaller period of time (1 or 2 days), which may increase recall and has similarities to a 24HR (a 1 day time period). Its scale is also more responsive than a typical FFQ frequency scale. However, it cannot be classified as a proper 24HR given that specific consumption is not recorded (see an explanation of a 24HR in the next section), nor a FFQ because of the atypical approach to the time period and scale.
In the past day, how many times did you consume the following foods and drinks?
Think about all the meals and snacks you ate. Also think about everything you had at home or outside the home. It is important that you report everything in the below categories that you had to eat or drink during this time. The past day refers to the previous day (from midnight to midnight). Please take your time.
| In the past day, how many times did you consume the following foods and drinks? | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10+ |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Meat (any type of meat, including beef, pork, chicken, turkey, fish, shellfish, or other meats) | |||||||||||
| Beef (hamburger, steak, roast beef, meatloaf, in stews / lasagna / pasta / pizza / etc.) | |||||||||||
| Pork (bacon, ham, pork chops, spare ribs, bacon bits, etc.) | |||||||||||
| Chicken (wings, nuggets, chicken breast, chicken sandwich, in soup, etc.) | |||||||||||
| Turkey (turkey dinner, turkey sandwich, turkey burger, turkey bacon, in soup, etc.) | |||||||||||
| Fish, not including shellfish (salmon, tuna, fish sticks, haddock, fish & chips, etc.) | |||||||||||
| Shellfish (lobster, shrimp, scallops, oyster, crab, etc.) | |||||||||||
| Other meats (duck, lamb/sheep, goat, bison, etc.) | |||||||||||
| Dairy (cheese, milk, yogurt, ice cream, butter, etc.) | |||||||||||
| Eggs (boiled, fried, omelet, in salad, in baked goods, etc.) |
24HR & ASA24
The NCI defines a 24-hour dietary recall (24HR) as:
a structured interview intended to capture detailed information about all foods and beverages (and possibly, dietary supplements) consumed by the respondent in the past 24 hours, most commonly, from midnight to midnight the previous day. […] This open-ended response structure is designed to prompt respondents to provide a comprehensive and detailed report of all foods and beverages consumed.
While 24HRs are commonly administered by a trained interviewer, a free tool has emerged to allow self-administration online: NCI’s Automated Self-Administered 24-Hour Recall (ASA24). Hundreds of Researchers have used ASA24 to date. If the study design will allow (considerations include respondent burden, respondent Internet access, cost, proficiency of analysts, etc.), we expect this method will produce high quality data.
NCI estimates that the average participant will need 22 minutes to complete the recall (without questions on supplements). Sample data files are provided to give Researchers a sense of the output they can expect. However, the predefined variables (e.g., M_MEAT and M_POULT) are not specific enough to answer all questions that movement Researchers may have. As such, we suggest using the outputs for MyPyramid equivalents (connected to food guidelines formerly used by the United States Department of Agriculture). Using for example the meat group from MyPyramid, it is possible to create groupings for the different types of meats. The only complication will be mixed dishes, which will require some estimates.
One downside of the ASA24 (and any 24HR) is that while only data on animal product consumption is needed, it assesses total dietary intake, which adds to respondent burden. However, we believe the benefits it brings in terms of accuracy outweighs this burden. Importantly, recalls alone cannot categorize individuals as vegans, vegetarians, pescetarians, meat reducers, etc.
Food Frequency Questionnaires
Food frequency questionnaires (FFQs) are a standard measurement tool from nutritional science, present less of a burden on respondents than 24HR instruments, and can be used to measure usual consumption. We see benefit to the format given below (and included on the complete list of recommended questions), however we hope upcoming research will clarify the benefits of going with longer vs. shorter frequency scales along with other design elements of these measures. It can be used with or without the columns for serving size. Including serving size allows more detailed calculations regarding total consumption to be made. Instructions for modifying the instrument and calculating consumption are given below the example.
FFQ Example – Without Serving Sizes
Thinking about your diet over the past three months, please select the responses that best describe how often you eat each type of food. Select only one response per row, even if different responses have been correct for different days or weeks.
| How often, in the past 3 months, did you eat the following? | never | Less than 1 time per MONTH | 1-3 times per MONTH | Once per WEEK | 2-4 times per WEEK | 5-6 times per WEEK | Once per DAY | 2-3 times per DAY | 4 or more times per DAY |
|---|---|---|---|---|---|---|---|---|---|
| Meat (any type of meat, including beef, pork, chicken, turkey, fish, shellfish, or other meats) | |||||||||
| Beef (hamburger, steak, roast beef, meatloaf, in stews / lasagna / pasta / pizza / etc.) | |||||||||
| Pork (bacon, ham, pork chops, spare ribs, bacon bits, etc.) | |||||||||
| Chicken (wings, nuggets, chicken breast, chicken sandwich, in soup, etc.) | |||||||||
| Turkey (turkey dinner, turkey sandwich, turkey burger, turkey bacon, in soup, etc.) | |||||||||
| Fish, not including shellfish (salmon, tuna, fish sticks, haddock, fish & chips, etc.) | |||||||||
| Shellfish (lobster, shrimp, scallops, oyster, crab, etc.) | |||||||||
| Other meats (duck, lamb/sheep, goat, bison, etc.) | |||||||||
| Dairy (cheese, milk, yogurt, ice cream, butter, etc.) | |||||||||
| Eggs (boiled, fried, omelet, in salad, in baked goods, etc.) |
FFQ Example – With Serving Sizes
Thinking about your diet over the past three months, please select the responses that best describe how often you eat each type of food and how much of it you eat at a time. Select only one frequency and one serving size per row, even if different responses have been correct for different days or weeks.
You may download either of the above tables in *.doc format.
Modifications
FFQs are flexible instruments that can easily be modified to address past or intended future diet. They can also be adjusted to be shorter and easier to complete or longer and more informative. Below, we suggest modifications that allow for customization while retaining the possibility of cross-comparison across studies using different FFQs.
Time Period
The example above addresses diet in the past three months. This time period was chosen to give a view of respondents’ present diet, with three months a long enough time period that brief and temporary changes (such as a vacation or illness) would not be likely to substantially alter responses. FFQs can be modified to ask about past diet by specifying time periods that are in the past. It is helpful to use specific dates in this case, so that respondents do not have to do mental calculation. For instance, “from January 1 to March 31, 2016” is more useful than “9 months to 1 year ago”.
FFQs could also address future or planned diet by using dates in the future, but note that people are not generally very reliable sources of information about their future habits. (For instance, people are much more likely to claim they will vote in an upcoming election than to vote in it, and we all have personal experience with times when we or others around us have not followed through on plans.) Measurements of respondents’ intentions with regard to dietary changes can provide interesting information about the persuasiveness of a message, but are not as telling as measurements of actual dietary change.
Of note, there can be both daily and seasonal effects on diet, which for some study designs will need to be taken into account.
Food List
The choice of food list plays a significant role in the information available from an FFQ, and many factors should be considered, including:
Length To increase response rates and quality, use the shortest list that meets the needs of the study. The lists we provide below are very short compared to lists used in nutritional and medical studies, which can contain hundreds of individual foods.
Balance of animal products and plant-based foods Survey responses can be affected by the perceived purpose of the survey. In theory, choosing a list that asks many more questions about animal products than about plant-based foods can indicate to the respondents what aspect of their diet the Researchers find interesting. Similarly, including other categories of food that have social or medical significance, like caffeinated beverages, may provide respondents with alternative possibilities to consider about the purpose of the survey, leading to less biased answers on questions about animal products.
The first iteration of our FFQ included “plant-based distractors,” a series of non-animal foods—e.g., fruits, vegetables, grains, beans, caffeinated drinks, sweeteners—that were included in the food list to help mask the purpose of the question in an effort to curb what we expected would be an under-reporting of animal products with a purely animal-based list. Since then, two unpublished studies have suggested that these distractors do not limit under-reporting of animal product consumption: 1) Do We Need to Disguise Veg Surveys? , and 2) a study on food lists and dietary interventions as biasing factors in animal product food frequency questionnaires. Based on these early findings, we do not recommend disguising FFQ food lists in intervention or non-intervention contexts. However, further (particularly published) research should be undertaken to confirm these preliminary findings.
Specificity of animal products listed Not all animal products are equally important to ask about specifically. Products which are consumed more frequently and which lead to greater suffering are more important to ask about than products which are consumed infrequently or lead to less suffering. Accordingly, our list places great emphasis on chicken consumption (based on its position in the list), since it is frequent and leads to the suffering of more individuals per serving than do many other animal products. We place less emphasis on infrequently consumed species such as rabbit, combining them into a single category (i.e., other meat). Our longest list (below) also separates cheese from the rest of the dairy category, since in the U.S., more milk is used to produce cheese than to produce any other dairy food.
Cultural relevance Our lists, including the memory aids, were developed with a broadly typical U.S. diet in mind. In adapting an FFQ for use in another country or with a U.S. population for which the foods listed are not especially common, culturally appropriate foods should be used to replace those not thought to be consumed frequently by the target population.
Order of items Respondents are most likely to answer questions accurately after they have become used to the format of the questionnaire, but before they have become bored. Items to which a correct response is particularly important (e.g., chicken) should be placed near but not at the beginning of the list.
Below we provide two food lists with memory aids, including the one used above and a longer list which would be more useful for producing precise calculations of the quantity of various animal products consumed.
To aid respondents, if more than one FFQ is present on a survey, each should use the same food list if possible.
Comprehensive List
Description: This list includes many subcategories of animal derived foods. The larger number of categories should help respondents more thoroughly recall their consumption habits. Additionally, separate categories for side and main dishes help in assigning accurate portion sizes and calculating the number of animals affected.
When to use this list: Use this list when a more precise calculation of consumption is required and when survey length is not a notable concern.
| Food | Long Memory Aid | Short Memory Aid |
|---|---|---|
| Cheese | processed (e.g. grilled cheese sandwich, cheeseburger), hard cheese, grated (e.g. on pizza or pasta), macaroni and cheese, in baked goods (e.g., cheese biscuit), etc. | in sandwich, on pizza/pasta, as hard cheese, etc. |
| Dairy main | milk/chocolate milk (e.g., as a beverage, on cereal), yogurt, ice cream, etc. | milk, yogurt, ice cream, etc. |
| Dairy side | butter, whip cream, sour cream, coffee cream, buttermilk, in baked goods, etc. | butter, cream, in baked goods, etc. |
| Dairy other | from non-cow sources, e.g., goats | from non-cow sources, e.g., goats |
| Chicken main | chicken wings, chicken breast, chicken nuggets, chicken strips, fried chicken, barbecue chicken, chicken deli sandwich, drumsticks, chicken burger, etc. | chicken breast, fried chicken, chicken deli sandwich, etc. |
| Chicken side | chicken noodle soup, chicken salad, chicken gravy, etc. | chicken soup, chicken salad, chicken gravy, etc. |
| Turkey main | turkey dinner, turkey burger, turkey sandwich, etc. | turkey dinner, turkey burger, turkey sandwich, etc. |
| Turkey side | turkey gravy, etc. | turkey gravy, etc. |
| Fish, not including shellfish | salmon, tuna, haddock, cod, herring, mackerel, sardines, anchovies, fish sticks, etc. | salmon, tuna, fish sticks, etc. |
| Shellfish | lobster, shrimp, scallops, oyster, crab, mussels, clams, etc. | lobster, shrimp, oyster, etc. |
| Pork main | bacon, ham, ham sandwich, pork chops, spare ribs, baby back ribs, pork tenderloin, etc. | ham, pork chops, ribs, etc. |
| Pork side | BLT sandwich, on pizza, bacon bits, pork and beans, pork fried rice, etc. | on pizza, bacon bits, in fried rice, etc. |
| Beef main | hamburger, steak, roast beef, meatloaf, pot roast, corned beef, beef short ribs, ground beef, etc. | hamburger, steak, roast beef, etc. |
| Beef side | beef gravy, beef stew, meatballs, on pizza, in tacos, in burritos, in lasagna, beef jerky, etc. | meatballs, on pizza, in tacos, etc. |
| Other beef/pork products | hotdogs, sausages, pepperoni, salami, bologna, etc. | hotdogs, sausages, pepperoni, etc. |
| Other meats | duck, lamb, goat, goose, rabbit, venison, bison, liver, etc. | duck, lamb, venison, etc. |
| Eggs main | scrambled, boiled, omelet, egg salad sandwich, quiches, etc. | scrambled, omelet, egg salad, etc. |
| Eggs side | in pasta, in sauces/dressings, in baked goods, etc. | in pasta, sauces, baked goods, etc. |
2. Short List
Description: This list is very short, and thus quicker and easier to administer than the comprehensive list. Enough categories are provided to avoid problems caused by basic definitional confusions (such as the idea that chicken is not meat), though it provides less support for memory than the comprehensive list, so recall is likely to be less thorough.
When to use this list: Use this list when the comprehensive list is inappropriate because respondents will not have time or attention to complete it.
| Food | Memory Aid |
|---|---|
| Meat | any type of meat, including beef, pork, chicken, turkey, fish, shellfish, or other meats |
| Beef | hamburger, steak, roast beef, meatloaf, in stews/lasagna/pasta/pizza, etc. |
| Pork | bacon, ham, pork chops, spare ribs, bacon bits, etc. |
| Chicken | wings, nuggets, chicken breast, chicken sandwich, in soup, etc. |
| Turkey | turkey dinner, turkey sandwich, turkey burger, turkey bacon, in soup, etc. |
| Fish, not including shellfish | salmon, tuna, fish sticks, haddock, fish & chips, etc. |
| Shellfish | lobster, shrimp, scallops, oyster, crab, etc. |
| Other meats | duck, lamb/sheep, goat, bison, etc. |
| Dairy | cheese, milk, yogurt, ice cream, butter, etc. |
| Eggs | boiled, fried, omelet, in salad, in baked goods, etc. |
3. Meat-only List
Description: This list is very short, and thus quicker and easier to administer than the comprehensive list. Enough categories are provided to avoid problems caused by basic definitional confusions (such as the idea that chicken is not meat), though it provides less support for memory than the comprehensive list, so recall is likely to be less thorough. It does not address dairy or egg consumption.
When to use this list: Use this list when you are only interested in vegetarianism or meat reduction behaviors, rather than veganism or all animal product consumption.
| Food | Memory Aid |
|---|---|
| Meat | any type of meat, including beef, pork, chicken, turkey, fish, shellfish, or other meats |
| Beef | hamburger, steak, roast beef, meatloaf, in stews/lasagna/pasta/pizza, etc. |
| Pork | bacon, ham, pork chops, spare ribs, bacon bits, etc. |
| Chicken | wings, nuggets, chicken breast, chicken sandwich, in soup, etc. |
| Turkey | turkey dinner, turkey sandwich, turkey burger, turkey bacon, in soup, etc. |
| Fish, not including shellfish | salmon, tuna, fish sticks, haddock, fish & chips, etc. |
| Shellfish | lobster, shrimp, scallops, oyster, crab, etc. |
| Other meats | duck, lamb/sheep, goat, bison, etc. |
Frequency Scale
A frequency scale should be chosen by balancing the precision of the information obtained with the increased difficulty for respondents when more response categories are available. We provide three scales, including the one used in the example above. The choice of frequency scale can affect the results received. Respondents often tend to choose responses towards the middle of the scale, so selecting a scale that fits with the food list to make common responses lie near the center of the scale is ideal. We always include “never” as a possible response, because of its significance in terms of respondent memory and to allow those administering a FFQ to identify respondents who are following a vegan or vegetarian diet.
To aid respondents, if more than one FFQ is present on a survey, each should use the same frequency scale if possible.
| 9-point scale (suitable for food lists with all animal products) | never | less than 1 time per MONTH | 1–3 times per MONTH | once per WEEK | 2-4 times per WEEK | 5-6 times per WEEK | Once per DAY | 2-3 times per DAY | 4 or more times per DAY |
| 9-point scale (suitable for food lists with meat only ) | never | Less than 1 time per MONTH | Once per MONTH | 2-3 times per MONTH | Once per WEEK | 2-4 times per WEEK | 5-6 times per WEEK | Once per DAY | 2 or more times per DAY |
| 7-point scale | never | less than 1 time per MONTH | 1-3 times per MONTH | 1-3 times per WEEK | 4-6 times per WEEK | once per DAY | 2 or more times per DAY | ||
| 5-point scale | never | less than 1 time per WEEK | 1-3 times per WEEK | 4-6 times per WEEK | 1 or more times per DAY |
Serving Size Scale
The inclusion of a serving size scale is optional, but allows for more precise calculation of the number of animals affected by a respondent’s diet. If serving sizes are included, they can be included parallel with the frequency scale, as in the first example above, or in a separate question following the other part of the FFQ. Serving sizes may be given in words, as in the examples we present, or with pictures, especially if the survey is administered on a computer. Because most people find serving sizes difficult to estimate from verbal descriptions, pictures may produce more accurate responses.
Use the items from the serving size list below that correspond to the food list on the survey. If adjustments are necessary to make the food list and memory aids culturally appropriate, they should also be considered for the example serving sizes given. If the serving size questions are presented separately from the frequency questions, keep the food list in the order used for the frequency questions.
For broad categories, an effort has been made to make serving sizes of different items in the category roughly interchangeable, in terms of the amount of animal product used.
| Standard serving size | 1 | 1/2 or less | 1 1/2 or more | |
|---|---|---|---|---|
| Cheese | 2 oz; 3 slices cheese; 1 slice pizza | |||
| Dairy main | 1 cup milk; 1 cup yogurt; 1 ½ cups ice cream | |||
| Dairy side | 1 tbsp butter or sour cream; 1 packet of creamer | |||
| Dairy other | 1 cup goat’s milk; 2 oz goat or sheep cheese | |||
| Chicken main | 3-4 oz; ½ large or 1 small breast; 2 drumsticks | |||
| Chicken side | 1 cup soup; ½ cup chicken salad; 2 tbsp gravy | |||
| Turkey main | 3-4 oz; 6-8 very thin slices; 1-3 thick slices | |||
| Turkey side | 2 tbsp gravy; 1 cup soup | |||
| Fish (not shellfish) | 3-4 oz; 1 can of tuna; 1 small or ½ medium fillet | |||
| Shellfish | 3-4 oz; 6 medium shrimp; 6-12 oysters | |||
| Pork main | 3-4 oz; 1 pork chop; 2 ribs; 3-4 slices bacon | |||
| Pork side | 1 piece pepperoni pizza; 1 cup pork fried rice | |||
| Beef main | 3-4 oz; ¼ lb burger; 3-6 slices roast beef | |||
| Beef side | 2 tbsp gravy; 1 meatball; 1 oz beef jerky | |||
| Other beef/pork | 3-4 oz; 1-2 hot dogs; 4 breakfast links | |||
| Other meats | 3-4 oz; a piece about the size of your palm | |||
| Eggs main | 1 egg; ⅓ cup scrambled eggs or egg salad | |||
| Eggs side | 2 tbsp dressing; 1 cup egg noodles; 1 slice cake | |||
| Dairy | 3 slices cheese; 1 cup milk; 1 cup yogurt | |||
| Eggs | 1 egg; ⅓ cup scrambled eggs or egg salad | |||
| Chicken | 3-4 oz; ½ large or 1 small breast; 2 drumsticks | |||
| Turkey | 3-4 oz; 6-8 very thin slices; 1-3 thick slices | |||
| Fish and shellfish | 3-4 oz; 1 can of tuna; 6 medium shrimp | |||
| Pork | 3-4 oz; 1 pork chop; 2 ribs; 3-4 slices bacon | |||
| Beef | 3-4 oz; ¼ lb burger; 3-6 slices roast beef | |||
| Other meat | 3-4 oz; a piece about the size of your palm |
Calculations
A food frequency questionnaire can be used without significant calculation, if it is only intended to determine whether respondents eat a certain type of food at all, although in this case it may simply be better to use the food list with a question such as “Which do you currently eat?” to reduce the burden on respondents. If a measure of overall consumption levels is required, calculation will be necessary.
Without Serving Sizes
If no serving size scale was included, the best justified calculation that can be made is the frequency of consumption. Ranges on the frequency scale can be converted into frequencies of consumption per day and averaged over the relevant group of respondents.
Our suggested daily frequencies for the categories on the frequency scales given above follow:1
| Reported frequency | never | less than 1 time per MONTH | 1–3 times per MONTH | once per WEEK | 2-4 times per WEEK | 5-6 times per WEEK | Once per DAY | 2-3 times per DAY | 4 or more times per DAY |
| Daily frequency | 0 | .016 | .066 | .142 | .427 | .784 | 1 | 2.5 | 4.5 |
| Reported frequency | never | Less than 1 time per MONTH | Once per MONTH | 2-3 times per MONTH | Once per WEEK | 2-4 times per WEEK | 5-6 times per WEEK | Once per DAY | 2 or more times per DAY |
| Daily frequency | 0 | .016 | .033 | .082 | .142 | .427 | .784 | 1 | 2.5 |
| Reported frequency | never | less than 1 time per MONTH | 1-3 times per MONTH | 1-3 times per WEEK | 4-6 times per WEEK | once per DAY | 2 or more times per DAY | ||
| Daily frequency | 0 | .016 | .066 | .285 | .712 | 1 | 2.5 | ||
| Reported frequency | never | less than once per WEEK | 1-3 times per WEEK | 4-6 times per WEEK | 1 or more times per DAY | ||||
| Daily frequency | 0 | .071 | .285 | .712 | 1.5 |
Once daily amounts are obtained for all respondents in the group, they may be averaged to produce an average daily amount. For example, if the sample consisted of only these two respondents, the overall average daily frequencies would be 1.106 servings of dairy per day (i.e., (1.5 + 0.712)/2), 0.178 servings of chicken per day (i.e. (0.285 + 0.071)/2), etc. Non-responses would be left out entirely, and the average taken over respondents who provided a response to the particular question.
With Serving Sizes
Including serving sizes allows an informed conversion to be made from daily frequency to ounces consumed daily, and from there to the number of animals affected. The first step in calculating the number of animals affected, therefore, is to compute the daily frequency for each respondent, as above. The daily amount (measured in servings) is this frequency, multiplied by a factor of 1, 0.5, or 1.5, depending on the serving size selected. If no selection is made for serving size, the response may be omitted, or the serving size may be taken to be 1; a choice should be made systematically for the entire data set.
Once daily amounts are obtained for all respondents in the group, they may be averaged to produce an average daily amount. For example, if the sample consisted of only these two respondents, the overall average daily amounts would be computed as 1.106 servings of dairy per day (i.e., 1.5 + 0.712/2), 0.195 servings of chicken per day (i.e. (0.106 + 0.285)/2), etc.
Average daily amounts can then be used to calculate numbers of animals affected per person per year, based on the standard serving sizes used. For instance, in the example above, 1.5 servings of dairy per day is about 12 oz of milk per day, or 274 lbs per year, which is about 1.5% of one dairy cow’s annual production. Similarly, .285 servings of chicken per day is about 1 oz of chicken per day, or 23 lbs per year, which is the meat from about 8 chickens. These calculations must be done separately for each category of animal product. Further adjustments should also be made for elasticity, to account for the response of economic systems to changes in supply or demand. While this can be done in a more complicated way, we have also seen some studies use a 50% elasticity factor for all products.
If serving size information is not solicited, conversions to animals affected may be made by assuming all respondents are using typical portion sizes, but there is a lack of data on what these sizes are, particularly for the broad categories of foods on our food lists, so this is expected to introduce significant potential for error into calculations.
Conclusion
We hope this page has given you the tools you need to select and use a dietary recall instrument for your survey. We encourage you to use the example questionnaires above or to modify them by choosing the food list, memory aids, and frequency scale that best suit your needs. Any questionnaire can be used with or without serving size information. If you have further questions about what questionnaire would be right for your purposes, or about how to modify a questionnaire for a substantially different audience, you can contact us or get more intensive help through Statistics Without Borders or Faunalytics.
Resources
Cade, J., Thompson, R., Burley, V., & Warm, D. (2002). Development, validation and utilisation of food-frequency questionnaires – A review. Public Health Nutrition, 5(4), 567-587.
Kristal, A. R., Baresford, S. A. A., & Lazovich, D. (1994). Assessing change in diet-intervention research. The American Journal of Clinical Nutrition, 59(1), Suppl: 185S–189S.
Mercy For Animals. (n.d.). Cruelty on screen: Do online videos of farmed animal cruelty change people’s diets and attitudes? Retrieved from http://www.mercyforanimals.org/impact-study
Molag, M. L., de Vries, J. H. M., Ockâe, M. C., Dagnelie, P. C., van den Brandt, P. A., Jansen, M. C. J. F., van Staveren, W. A., & van’t Veer, P. (2007). Design characteristics of food frequency questionnaires in relation to their validity. American Journal of Epidemiology, 166(12), 1468-1478.
National Cancer Institute. (n.d.). 24-hour dietary recall (24HR) at a glance. Retrieved from https://dietassessmentprimer.cancer.gov/profiles/recall/
National Cancer Institute. (n.d.). Evaluating the effect of an intervention on diet. Retrieved from https://dietassessmentprimer.cancer.gov/approach/intervention.html
National Cancer Institute. (n.d.). Summary tables: recommendations on potential approaches to dietary assessment for different research objectives requiring group-level estimate. Retrieved from https://dietassessmentprimer.cancer.gov/approach/table.html
NutritionQuest: More information about assessment tools.
Thomson, C. A., Giuliano, A., Rock, C. L., Ritenbaugh, C. K., Flatt, S. W., Faerber, S. , . . . Marshall, J. R. (2003). Measuring dietary change in a diet intervention trial: Comparing food frequency questionnaire and dietary recalls. American Journal of Epidemiology, 157(8), 754–762.